312 CHEMICAL ENGINEERING so small that the dust particles will not rise properly. If there be a fan and on one side it is pulling and on the other pressing against the resistances necessary in filtration or other separation means there must either be a serious loss in pressure or an unnecessarily large fan must be employed. All this is avoided by merely having a dust chamber of ample cross-section and as free as possible from bends and turns. The power losses in bends and contractions are well analyzed by lies.1 Where only the finest part of the dust is of commercial value a tortuous or constricted chamber will create vortices and there will be settlement by centrifugal force with removal of fines at the wrong point. The best form of dust chamber is one of comparatively great length compared with the other two dimensions and it has the additional advantage that hoppers can be more cheaply placed in the bottom and with less loss of head room. Also since the bulk of the material which will settle promptly will be found near the entry of the chamber less material has to be discharged in the long chamber as the dust at the far end needs only occasional removal. There is also a more practical gradation in size of particle from entry to exit end.2 "Water in Dust Settlement.—In nearly every case where water can be used it has a beneficial effect. It is of much assistance in the Cottrell process of dust settlement to be described later. The more finely divided the water, the less need be the amount used but that the effect is somewhat independent of the quantity, that is that a large quantity in bulk will do effective work is shown by the experiments of lies.3 Jets of water were allowed to flow down through a tin box with slightly inclined alternate shelves so as to give the water a tortuous course. From 90 to 95 per cent of the smoke and fume was collected from the gas ascending the box.4 Humidifying can be accomplished by nozzles or by centrifugal apparatus, which whirls the water into an impalpable mist. Separation by Centrifugal Force.—The "Draco" separator is illustrated in Fig. 32. The figure also illustrates a water attachment which has much merit on material which will not be injured by water. Separators resembling the dry portion of this separator are manufactured by all the makers of grain cleaning machinery. The water attachment which gives atomized water by a series of spraying nozzles cannot be used successfully on materials which have no affinity for water such as coal dust, graphite dust, soapstone. One of the suggested uses for it is hydrating the dust which arise from lime hydrators. 1 "Lead Smelting," p. 146 et seq., John Wiley & Sons, New York, 1902. 2 It has been proposed to build whole dust chambers out of extremely porous brick and let the chamber have no vent except through the porous walls. The difficulty appeared to be that eventually the pores fill up and cannot be freed.—EDITOR. 8 Op. cit. * Whether these results can be duplicated on a larger scale is open to some question. The experience of silver refineries with scrubbers where the water was not atomized has been that much valuable material goes by them. Despite its highly hygroscopic character, phosphorus pentoxide can be carried in a current of air through a number of wash bottles. In several scrubbers in which a multiplicity of 4-in. pipes carried the gases from a smoke-box into which they were forced by pressure down through water (the gases bubbling from the end of the pipes) the recovery was extremely poor. According to lies a weak vinegar solution would have been more effective than pure water. A good experimental scrubber that I once saw consisted of a flattened cone containing a multiplicity of small holes and immersed in water. The gas was delivered at the top of the cone under pressure and then passed out the small holes into the water. Scarcely a trace of fume passed it, but this was a small installation handing only about 75 cu. ft. of gas per minute. That the ordinary scrubber is ineffective can be shown by drawing a known quantity of the effluent gas through absorbent cotton. The teachings of the colloidal chemists would lead to the belief that no scrubber could be successful except where the water was highly subdivided, and the success of the Feld scrubber is a brilliant testimony to the soundness of their theories.—EDITOR.